CN103163534A - Adaptive noise bandwidth carrier loop tracking method - Google Patents

Abstract

The invention provides an adaptive noise bandwidth carrier loop tracking method. The method sequentially includes the steps: stripping a carrier; accurately stripping pseudo codes; performing coherent integration; detecting a mean square error of in-phase branch coherent accumulated values; adaptively adjusting noise bandwidths; detecting phases; filtering a loop; and feeding back and controlling a carrier NCO (numerically controlled oscillator). The normalized mean square error of the coherent accumulated values of in-phase channels serves as a judgment threshold to adaptively adjust the noise bandwidth of a carrier tracking loop, the influence of dynamic stress change and loop noise on loop convergence conditions can be effectively balanced in real time, and loop convergence time and accuracy are effectively improved.

Description

A kind of carrier loop tracking of adaptive noise bandwidth

Technical field

The present invention relates to a kind of adaptive noise bandwidth carrier loop tracks method, belong to the satellite navigation field.

Background technology

Global Navigation Satellite System (GNSS) can round-the-clockly provide accurate PVT(position, speed, time in the world) information, have very wide purposes and development prospect.The GNSS receiver is the equipment of receiving satellite signal, and the satellite-signal that its utilization receives obtains necessary Navigation and localization information and observed quantity, and processes through data and finally realize Navigation and localization.In once complete navigator fix process, at first obtain the Doppler shift of satellite-signal and the initial value of code phase by catching operation, then, enter into carrier tracking loop, further determine the code phase of carrier frequency and spreading code, simultaneously satellite-signal is followed the tracks of continuously accurately, and then demodulated required navigation data, and obtain final position and velocity information by operations such as extracting ephemeris, its computation of pseudoranges, satellite position and resolve, the GNSS receiver location resolves.

At present, have much about the research of carrier tracking loop, but conclusion is all to obtain under the prerequisite that carrier tracking loop noise bandwidth does not change with the input signal signal to noise ratio (S/N ratio) mostly.Putting before this, improving the most direct method of tracking sensitivity and improve exactly the input signal signal to noise ratio (S/N ratio), namely increasing the coherent accumulation time.But the growth of coherence time can cause again the ability of carrier tracking loop reply high dynamic environment to weaken, and the phenomenon of carrier tracking loop losing lock even occurs.

In document 1 " Rapid Parallel GPS Signal Acquisition[ION GPS; Salt Lake City; UT; September19-22; 2000] ", the phaselocked loop of design has adopted the two different loop parameters of cover, respectively corresponding two loop filters that noise bandwidth is different, after the receiver signal passage enters tracking phase, first allowing phaselocked loop adopt bandwidth is the traction wave filter of 110Hz, and allow it move 30ms, phase tracking error is reduced rapidly, and then switch to the tracking filter that bandwidth is 15Hz.But such scheme just with tracking time as the state switching condition, can not be strong and weak different with dynamic stress and in real time, adaptively loop is adjusted along with the noise transformation situation, thus can't be operated under high environment dynamically.

Summary of the invention

In order to overcome the deficiencies in the prior art, the invention provides a kind of adaptive noise bandwidth carrier track loop method, with in-phase branch coherent accumulation value I _PStatistical variance as decision threshold, adjust adaptively carrier tracking loop noise bandwidth, improve to follow the tracks of the tracking performance of carrier tracking loop under high dynamic environment.

The technical solution adopted for the present invention to solve the technical problems comprises the following steps:

1) the two-way local carrier difference mixing after the process mapping that will export from digital medium-frequency signal and the carrier wave NCO of tracking front-end processing system input forms homophase, orthogonal double channels signal;

2) do relevantly in homophase, orthogonal double channels signal input code tracking loop front end respectively,, hysteresis branch road pseudo code correlation device leading, instant in this locality, be output as six road correlated results; Described lead and lag branch road is, hysteresis half chip more leading than instant branch road respectively;

3) coherent integration that the two-way correlated results of instant branch road correlator being exported is respectively 1ms-10ms adds up, and exports respectively two-way coherent accumulation value, then removes each deposit unit of integrator, then carries out the integration of next period, so repeats constantly.

4) in-phase branch coherent accumulation value mean square deviation detects and the adjustment of noise bandwidth self-adaptation, and detailed process is as follows:

A) receive homophase passage coherent accumulation value I _PAnd real-time storage, with storing value carry out carrying out successively after the N point sampling normalization mean square deviation in step b) is calculated and step c) in the thresholding judgement, at this section in the period, with the I of reception in real time _PValue directly transfers to the carrier wave phase detector, participates in calculating optimum noise bandwidth in the PLL loop as loop parameter with optimum noise bandwidth in the PLL loop:

${\left(B_n\right)}_{\mathrm{PLL}-\mathrm{optimum}}=\sqrt[7]{\frac{{\left(\left(2n\right)\frac{a}{3}\mathrm{\&Delta;f}\right)}^2}{\frac{1}{c/n_0}(1+\frac{1}{2\mathrm{Tc}/n_0}){\left(\frac{{\mathrm{\&lambda;}}_{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA00002903081500012.png"\; state="\{\{state\}\}">L</figure-callout>}}}{2\mathrm{\&pi;}}\right)}^2}}$

Wherein, a=B _nω, ω are loop undamped oscillation frequency, and Δ f is initial frequency difference, c/n ₀Be the input signal carrier-to-noise ratio, T is the coherent integration time, λ _LBe carrier signal wavelength;

B) establish the I that needs statistics _PSampled value is respectively { I _pi, wherein, i=0,1 ..., N-1 is according to I _piSign set respectively two the set { I ⁺And { I ^-, satisfy following formula

$I_{\mathrm{pi}}\&Element;\left\{\begin{array}{cc}{I}^{+},& I_{\mathrm{pi}}\&GreaterEqual;0\\ I^{-},& I_{\mathrm{pi}}<0\end{array}\right.$

Further establish

With

Be set { I in current statistics N point ⁺And { I ^-In the number of element, define current N point sampling normalization meansquaredeviationσ ²For:

${\mathrm{\&sigma;}}^2=\frac{\mathrm{Var}\left(\right\{I^{+}\left\}\right)\&times;N_{I^{+}}+\mathrm{Var}\left(\right\{I^{-}\left\}\right)\&times;N_{I^{-}}}{E\left(\right\{I_{\mathrm{pi}}^2\left\}\right)\&times;N}$

Wherein,

$\mathrm{Var}\left(\right\{I^{+}\left\}\right)=\frac{1}{N_{I^{+}}}\underset{0}{\overset{N_{I^{+}}-1}{\mathrm{\&Sigma;}}}{(I_i^{+}-E\left(\right\{I^{+}\left\}\right))}^2$

$E\left(\right\{I^{+}\left\}\right)=\frac{1}{N_{I^{+}}}\underset{0}{\overset{N_{I^{+}}}{\mathrm{\&Sigma;}}}{I}_i^{+}$

$\mathrm{Var}\left(\right\{I^{-}\left\}\right)=\frac{1}{N_{I^{-}}}\underset{0}{\overset{N_{I^{-}}-1}{\mathrm{\&Sigma;}}}{(I_i^{-}-E\left(\right\{I^{-}\left\}\right))}^2$

$E\left(\right\{I^{-}\left\}\right)=\frac{1}{N_{I^{-}}}\underset{0}{\overset{N_{I^{-}}}{\mathrm{\&Sigma;}}}{I}_i^{-}$

$E\left(\right\{I_{\mathrm{pi}}^2\left\}\right)=\frac{1}{N}\underset{0}{\overset{N-1}{\mathrm{\&Sigma;}}}{I}_{\mathrm{pi}}$

C) the normalization mean square deviation that calculates in previous step and the thresholding of setting are compared: if in threshold range, the state when catching rear initial the tracking is identical, still selects the optimum noise bandwidth that is calculated by the minimum loop tracking error; If greater than the thresholding upper limit, increase noise bandwidth, change with the reply dynamic stress; If less than the thresholding lower limit, the noise decrease bandwidth, make loop-locking more stable;

5) will be through step 4) homophase, the orthogonal channel coherent accumulation value incoming carrier phase detector of the real-time storage processed, its output is digital medium-frequency signal and the phase error of local carrier or the function of phase error of carrier tracking loop input;

6) phase error of carrier wave phase detector output or the function of phase error are input to loop filter, with step 4) in the self-adaptation that the obtains noise bandwidth after adjusting participate in the calculating loop filter parameters;

7) output of loop filter feeds back to carrier wave NCO, with corresponding increase or reduce the local carrier frequency that carrier wave NCO produces, so far complete line loop one time, after skip back to step 1) re-treatment, until the phase error of carrier wave phase detector output when being zero, is judged the loop stability convergence.

The invention has the beneficial effects as follows: the present invention adjusts carrier tracking loop noise bandwidth with the normalization mean square deviation of homophase passage coherent accumulation value adaptively as the judgement thresholding, can be in real time, the dynamic STRESS VARIATION of balance and loop noise on the impact of loop convergence situation, effectively improve loop convergence time and convergence precision performance preferably.

Description of drawings

Fig. 1 is adaptive noise bandwidth carrier track loop method flow diagram;

Fig. 2 is different noise bandwidth scheme IP value normalization mean square deviation comparison diagrams;

Fig. 3 is that adaptive noise bandwidth scheme loop is adjusted situation map;

Fig. 4 follows the tracks of different phase IP value normalization mean square deviation situation of change figure.

Embodiment

The present invention is further described below in conjunction with drawings and Examples.

The present invention proposes a kind of carrier tracking loop path method of adaptive noise bandwidth, as shown in Figure 1, comprises the following steps:

1) carrier wave is peeled off: will be from following the tracks of digital medium-frequency signal and the carrier wave NCO(numerical controlled oscillator of front-end processing system input) output through the two-way local carrier mixing respectively after the sin/cos mapping, formation homophase, orthogonal double channels signal.

2) pseudo-code is accurately peeled off: homophase, orthogonal double channels signal be input code tracking loop front end respectively, do relevantly in leading in this locality, instant, hysteresis branch road (lead and lag branch road respectively more leading than instant branch road, half chip lags behind) pseudo code correlation device, be output as six road correlated results.Be simplified model, only get the two-way correlated results of instant branch road correlator and output thereof herein and analyze, and think and accurately to peel off pseudo-code after relevant.

3) coherent integration: the coherent integration of the two-way correlated results of instant branch road correlator output being done respectively certain hour (being generally 1ms-10ms) is cumulative, export respectively two-way coherent accumulation value, then remove each deposit unit of integrator, then carry out the integration of next period, so repeat constantly.

4) detection of in-phase branch coherent accumulation value mean square deviation and noise bandwidth Adaptive adjusting algorithm are key component of the present invention, and detailed process is as follows:

A) receive homophase passage coherent accumulation value I _PAnd real-time storage, storing value is carried out the N(span selected according to system's real time processing accuracy) carry out successively that normalization mean square deviation in step b) is calculated after point sampling and step c) in thresholding judge.Because calculating, sampling and normalization mean square deviation all can not carry out in real time, so at this section in the period, with the I of reception in real time _PValue directly transfers to the carrier wave phase detector, participates in calculating as loop parameter with the initial optimum noise bandwidth that is calculated by the minimum loop tracking error that proposes in document 2.Document 2 " carrier track that combines based on FLL/PLL under dynamic environment [playing arrow and guidance journal, the 24th the 1st phase of volume, 171-173] " proposes the poor hour correspondence of loop tracks mean square of error the optimal loop bandwidth, and tries to achieve:

${\left(B_n\right)}_{\mathrm{PLL}-\mathrm{optimum}}=\sqrt[7]{\frac{{\left(\left(2n\right)\frac{a}{3}\mathrm{\&Delta;f}\right)}^2}{\frac{1}{c/n_0}(1+\frac{1}{2\mathrm{Tc}/n_0}){\left(\frac{{\mathrm{\&lambda;}}_{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA00002903081500012.png"\; state="\{\{state\}\}">L</figure-callout>}}}{2\mathrm{\&pi;}}\right)}^2}}---\left(1\right)$

Wherein, (B _n) _PLLoptimumOptimum noise bandwidth in the PLL loop, a=B _nω, ω are loop undamped oscillation frequency, and Δ f is initial frequency difference, c/n ₀Be the input signal carrier-to-noise ratio, T is the coherent integration time, λ _LBe carrier signal wavelength.

B) the normalization mean square deviation is calculated: the I that need to suppose statistics _PCounting of sampled value is the N point, and sampled value is respectively { I _pi, wherein, i=0,1 ..., N-1.According to I _piSign set respectively two the set { I ⁺And { I ^-, satisfy following formula

$I_{\mathrm{pi}}\&Element;\left\{\begin{array}{cc}{I}^{+},& I_{\mathrm{pi}}\&GreaterEqual;0\\ I^{-},& I_{\mathrm{pi}}<0\end{array}---\left(2\right)\right.$

Further suppose

With Be set { I in current statistics N point ⁺And { I ^-In the number of element, obviously, Thus, the current N point sampling normalization meansquaredeviationσ of definition ²For:

${\mathrm{\&sigma;}}^2=\frac{\mathrm{Var}\left(\right\{I^{+}\left\}\right)\&times;N_{I^{+}}+\mathrm{Var}\left(\right\{I^{-}\left\}\right)\&times;N_{I^{-}}}{E\left(\right\{I_{\mathrm{pi}}^2\left\}\right)\&times;N}---\left(3\right)$

Wherein,

$\mathrm{Var}\left(\right\{I^{+}\left\}\right)=\frac{1}{N_{I^{+}}}\underset{0}{\overset{N_{I^{+}}-1}{\mathrm{\&Sigma;}}}{(I_i^{+}-E\left(\right\{I^{+}\left\}\right))}^2---\left(4\right)$

$E\left(\right\{I^{+}\left\}\right)=\frac{1}{N_{I^{+}}}\underset{0}{\overset{N_{I^{+}}}{\mathrm{\&Sigma;}}}{I}_i^{+}---\left(5\right)$

$\mathrm{Var}\left(\right\{I^{-}\left\}\right)=\frac{1}{N_{I^{-}}}\underset{0}{\overset{N_{I^{-}}-1}{\mathrm{\&Sigma;}}}{(I_i^{-}-E\left(\right\{I^{-}\left\}\right))}^2---\left(6\right)$

$E\left(\right\{I^{-}\left\}\right)=\frac{1}{N_{I^{-}}}\underset{0}{\overset{N_{I^{-}}}{\mathrm{\&Sigma;}}}{I}_i^{-}---\left(7\right)$

$E\left(\right\{I_{\mathrm{pi}}^2\left\}\right)=\frac{1}{N}\underset{0}{\overset{N-1}{\mathrm{\&Sigma;}}}{I}_{\mathrm{pi}}---\left(8\right)$

C) thresholding judgement: the thresholding (the thresholding span is selected according to system's real time processing accuracy) that system in the normalization mean square deviation that calculates in previous step and this step has been set compares: if in threshold range, the state when catching rear initial follow the tracks of is identical, still selects the optimum noise bandwidth (as shown in Equation 1) that is calculated by the minimum loop tracking error; If greater than the thresholding upper limit, increase noise bandwidth, change with the reply dynamic stress; If less than the thresholding lower limit, the noise decrease bandwidth, make loop-locking more stable.Each adjustment amount viewing system real time processing accuracy of noise bandwidth and deciding.In practical application, under the Complex Noise background, also can set up the threshold value training module before this step thresholding judgement, change according to noise and set different threshold level, go out the threshold value of suitable current noise situations by real-time detection and feedback training.

Homophase, the orthogonal channel coherent accumulation value incoming carrier phase detector of the real-time storage of 5) phase demodulation: will be through step 4) processing, its output are the digital medium-frequency signal of carrier tracking loop input and the phase error (or function of phase error) of local carrier.

6) loop filtering: the phase error (or function of phase error) of carrier wave phase detector output is input to loop filter, with step 4) in the self-adaptation that the obtains noise bandwidth after adjusting participate in the calculating loop filter parameters.

7) FEEDBACK CONTROL carrier wave NCO: the output of loop filter feeds back to carrier wave NCO, with corresponding increase or reduce the local carrier frequency that carrier wave NCO produces, so far complete line loop one time, after skip back to step 1) re-treatment, until the phase error of carrier wave phase detector output when being zero, is judged the loop stability convergence.

Adopt the actual measurement gps signal in the present embodiment, the digital medium-frequency signal of input is 15.58MHz, and sample frequency is 62MHz, the Doppler shift after catching is controlled at ± 250Hz within, burst dynamic stress acceleration be changed to 2g, g is acceleration of gravity, and the local sine and cosine look-up table stores of the carrier wave NCO degree of depth is 512, and data bit width is 8, the coherent integration time is 1ms, the threshold range lower limit is made as 1.0e-4, and the upper limit is made as 1.0e-3, needs the I of statistics _PThe points N of sampled value is 25.

In accompanying drawing 1, I ₁, Q ₁Signal is respectively homophase, the orthogonal double channels signal of the digital medium-frequency signal of input and the local carrier mixing formation that carrier wave NCO produces; I ₂, Q ₂Be I ₁, Q ₁The result that the instant branch road of signal and code tracking loop (P branch road) correlator carries out related calculation and exports; To two-way correlated results I ₂, Q ₂Do respectively coherent integration and add up and clear operation, corresponding output two-way coherent accumulation value I _P, Q _P

Accompanying drawing 2 is depicted as according to the setting of the present embodiment parameter, adopts respectively steady noise bandwidth 10Hz, 60Hz, fixing optimum noise bandwidth 25Hz and based on I _PWhen being worth the adaptive noise bandwidth of normalization mean square deviation thresholding, I _PValue normalization mean square deviation situation of change.As can be seen, along with noise bandwidth reduces, the I during loop convergence _PValue normalization mean square deviation reduces gradually; Data acquisition sample value 5N place (corresponding tracking time 125ms place), I _PValue normalization mean square deviation is less than thresholding lower limit 1.0e-4, and this place is adjusted downward to 20Hz with noise bandwidth by initial 25Hz; (corresponding tracking time 500ms) detects I at sampled value 20N place _PValue normalization mean square deviation is lowered noise bandwidth again less than 0.1 times of the initial threshold lower limit.Accompanying drawing 3 is depicted as with accompanying drawing 2 identical parameters and arranges down, and during the adaptive noise bandwidth condition, corresponding loop is adjusted situation.Can see the adjustment along with twice noise bandwidth in figure, frequency offset tends towards stability gradually, makes it more to adapt to current loop noise situation, has improved well the loop convergence precision.

It is that other parameters of 2g(arrange identical with embodiment that following table is depicted as burst dynamic stress acceleration) time, four kinds of noise bandwidths (steady noise bandwidth 10Hz, fixedly optimum noise bandwidth 25Hz, steady noise bandwidth 60Hz and based on I selected respectively _PThe adaptive noise bandwidth of value normalization mean square deviation thresholding), I after loop convergence time, loop stability convergence first _PAfter value mean square deviation and burst dynamic stress, loop is restrained the contrast of situation again.Analytical data is bright, adaptive noise bandwidth scheme preferably balance convergence time and convergence precision, integral body has been improved loop performance.

Loop performance contrast under the different noise bandwidths of table 1 arrange

Claims (1)

1. the carrier loop tracking of an adaptive noise bandwidth, is characterized in that comprising the steps:

1) the two-way local carrier difference mixing after the process mapping that will export from digital medium-frequency signal and the carrier wave NCO of tracking front-end processing system input forms homophase, orthogonal double channels signal;

2) do relevantly in homophase, orthogonal double channels signal input code tracking loop front end respectively,, hysteresis branch road pseudo code correlation device leading, instant in this locality, be output as six road correlated results; Described lead and lag branch road is, hysteresis half chip more leading than instant branch road respectively;

3) coherent integration that the two-way correlated results of instant branch road correlator being exported is respectively 1ms-10ms adds up, and exports respectively two-way coherent accumulation value, then removes each deposit unit of integrator, then carries out the integration of next period, so repeats constantly.

4) in-phase branch coherent accumulation value mean square deviation detects and the adjustment of noise bandwidth self-adaptation, and detailed process is as follows:

A) receive homophase passage coherent accumulation value I _PAnd real-time storage, with storing value carry out carrying out successively after the N point sampling normalization mean square deviation in step b) is calculated and step c) in the thresholding judgement, at this section in the period, with the I of reception in real time _PValue directly transfers to the carrier wave phase detector, participates in calculating optimum noise bandwidth in the PLL loop as loop parameter with optimum noise bandwidth in the PLL loop:

${\left(B_n\right)}_{\mathrm{PLL}-\mathrm{optimum}}=\sqrt[7]{\frac{{\left(\left(2n\right)\frac{a}{3}\mathrm{\&Delta;f}\right)}^2}{\frac{1}{c/n_0}(1+\frac{1}{2\mathrm{Tc}/n_0}){\left(\frac{{\mathrm{\&lambda;}}_L}{2\mathrm{\&pi;}}\right)}^2}}$

Wherein, a=B _nω, ω are loop undamped oscillation frequency, and Δ f is initial frequency difference, c/n ₀Be the input signal carrier-to-noise ratio, T is the coherent integration time, λ _LBe carrier signal wavelength;

B) establish the I that needs statistics _PSampled value is respectively { I _pi, wherein, i=0,1 ..., N-1 is according to I _piSign set respectively two the set { I ⁺And { I ^-, satisfy following formula

$I_{\mathrm{pi}}\&Element;\left\{\begin{array}{cc}{I}^{+},& I_{\mathrm{pi}}\&GreaterEqual;0\\ I^{-},& I_{\mathrm{pi}}<0\end{array}\right.$

Further establish

With

Be set { I in current statistics N point ⁺And { I ^-In the number of element, define current N point sampling normalization meansquaredeviationσ ²For:

${\mathrm{\&sigma;}}^2=\frac{\mathrm{Var}\left(\right\{I^{+}\left\}\right)\&times;N_{I^{+}}+\mathrm{Var}\left(\right\{I^{-}\left\}\right)\&times;N_{I^{-}}}{E\left(\right\{I_{\mathrm{pi}}^2\left\}\right)\&times;N}$

Wherein,

$\mathrm{Var}\left(\right\{I^{+}\left\}\right)=\frac{1}{N_{I^{+}}}\underset{0}{\overset{N_{I^{+}}-1}{\mathrm{\&Sigma;}}}{(I_i^{+}-E\left(\right\{I^{+}\left\}\right))}^2$

$E\left(\right\{I^{+}\left\}\right)=\frac{1}{N_{I^{+}}}\underset{0}{\overset{N_{I^{+}}}{\mathrm{\&Sigma;}}}{I}_i^{+}$

$\mathrm{Var}\left(\right\{I^{-}\left\}\right)=\frac{1}{N_{I^{-}}}\underset{0}{\overset{N_{I^{-}}-1}{\mathrm{\&Sigma;}}}{(I_i^{-}-E\left(\right\{I^{-}\left\}\right))}^2$

$E\left(\right\{I^{-}\left\}\right)=\frac{1}{N_{I^{-}}}\underset{0}{\overset{N_{I^{-}}}{\mathrm{\&Sigma;}}}{I}_i^{-}$

$E\left(\right\{I_{\mathrm{pi}}^2\left\}\right)=\frac{1}{N}\underset{0}{\overset{N-1}{\mathrm{\&Sigma;}}}{I}_{\mathrm{pi}}$

C) the normalization mean square deviation that calculates in previous step and the thresholding of setting are compared: if in threshold range, the state when catching rear initial the tracking is identical, still selects the optimum noise bandwidth that is calculated by the minimum loop tracking error; If greater than the thresholding upper limit, increase noise bandwidth, change with the reply dynamic stress; If less than the thresholding lower limit, the noise decrease bandwidth, make loop-locking more stable;

5) will be through step 4) homophase, the orthogonal channel coherent accumulation value incoming carrier phase detector of the real-time storage processed, its output is digital medium-frequency signal and the phase error of local carrier or the function of phase error of carrier tracking loop input;

6) phase error of carrier wave phase detector output or the function of phase error are input to loop filter, with step 4) in the self-adaptation that the obtains noise bandwidth after adjusting participate in the calculating loop filter parameters;

7) output of loop filter feeds back to carrier wave NCO, with corresponding increase or reduce the local carrier frequency that carrier wave NCO produces, so far complete line loop one time, after skip back to step 1) re-treatment, until the phase error of carrier wave phase detector output when being zero, is judged the loop stability convergence.

Images